errors occur, causing the encoding windows to crash one after another.
Steps to Reproduce:
Prepare the necessary number of folders, each containing 30 full-color images with dimensions 2000×3500 pixels. (Large full-color images that take a long time to process are acceptable.)
Launch approximately three multi-encoding windows simultaneously.
Use all multi-threading options, and set jxl to lossless.
Set each encoding window with a folder containing images, and start converting them simultaneously.
System Specifications:
Operating System: Windows 11 24H2
CPU: Ryzen 5950X
Memory: 32GB
1.8.5 When running multiple instances of jxl file multi-encoding, ...
Moderators: helmut, XnTriq, xnview, Dreamer
Re: 1.8.5 When running multiple instances of jxl file multi-encoding, ...
are you able to reproduce each time?
Pierre.
- BloodySword
- Posts: 42
- Joined: Sun Mar 28, 2010 11:19 am
Re: 1.8.5 When running multiple instances of jxl file multi-encoding, ...
JPEG-XL multi threaded processing can take a lot of RAM, especially at higher resolutions.
I don't think this is an issue with XnView MP.
Try to use the command line tool cjxl.exe instead and try the same processing via command line. Check your memory usage when processing. If the same issue happens here, then it's nothing to do with XNView.
Possible troubleshooting steps:
- Make sure your pagefile has a size of at least 50% of your physical memory size.
- Run less conversions at the same time
- Specify fewer threads per file
I don't think this is an issue with XnView MP.
Try to use the command line tool cjxl.exe instead and try the same processing via command line. Check your memory usage when processing. If the same issue happens here, then it's nothing to do with XNView.
Possible troubleshooting steps:
- Make sure your pagefile has a size of at least 50% of your physical memory size.
- Run less conversions at the same time
- Specify fewer threads per file
Re: 1.8.5 When running multiple instances of jxl file multi-encoding, ...
My JPEG-XL tests also show, that big pictures use very much RAM for compression.
16 GB RAM is okay, but sometimes you need more ram especially for multi threading (2 or more cores).
A fast GUI and tools is the xl-converter. It also contains the jpegxl-encoder
and other tools.
Maybe it has a better RAM usage and controlling.
https://codepoems.eu/xl-converter
16 GB RAM is okay, but sometimes you need more ram especially for multi threading (2 or more cores).
A fast GUI and tools is the xl-converter. It also contains the jpegxl-encoder
and other tools.
Maybe it has a better RAM usage and controlling.
https://codepoems.eu/xl-converter
PC: Intel 8700k + Asus Z370-F + 16 GB RAM G.Skill + Asus RTX 3050 OC + Samsung SSD
OS: Windows 10 Pro 64bit 22H2 |
GFX: XnViewMP 1.8.7 | XN-View 2.52.0 | XnConvert 1.104.0 | Adobe Photoshop Elements 2024 | Elements XXL 11 | Paint.Net 5.1.7
OS: Windows 10 Pro 64bit 22H2 |
GFX: XnViewMP 1.8.7 | XN-View 2.52.0 | XnConvert 1.104.0 | Adobe Photoshop Elements 2024 | Elements XXL 11 | Paint.Net 5.1.7
Re: 1.8.5 When running multiple instances of jxl file multi-encoding, ...
Always reproducible in 1.8.5.
Mitigated in 1.8.6.
XncConvert7.210 operates stably.
Benchmarking cjxl.exe is difficult, requiring ideas and ingenuity. Thanks for the GUI. A GUI with structural understanding is needed to determine if an error is JPEG XL-related. (Command-line batch conversion is estimated to be limited to around 60 files at a time.)
Mitigated in 1.8.6.
XncConvert7.210 operates stably.
Benchmarking cjxl.exe is difficult, requiring ideas and ingenuity. Thanks for the GUI. A GUI with structural understanding is needed to determine if an error is JPEG XL-related. (Command-line batch conversion is estimated to be limited to around 60 files at a time.)